Hawke64

I think I had seen the other Dancing Dragon games before purchasing SoW, but they seemed like more generic RPGMaker games and I quite strongly dislike the JRPG combat in them. There are some exceptions (Crystal Tactics and Trans Theft Horso), but most go with the defaults. Could you please let me know what you liked and disliked about the Dancing Dragon titles?

Got the Soul Reaver remake. It looks like I imagined the original. I have little recollection of the combat and the story (got cooler than the boss => got kicked into a whirlpool => must get revenge), so it is nice to rediscover them. The lack of the thumb mouse button support is less nice, but generally fine. Regarding updates, I suppose my main issue is that now purchasing a game on release is effectively being a paying beta-tester, while each and every patch breaks things and, in the case of Steam, prevents you from playing. Mind, previously, what was broken stayed broken, so you could end up with a non-functioning or incredibly buggy game, but the current approach encourages the developers to release MVP and patch later. Thus, unless you really know and love the particular developer, waiting for at least a year or two usually provides a better experience. On the other hand, e.g., the Eternal Strands devs removed the DX11 support in one of the patches and rolling back on anything but Steam is impossible. That is to say, the experience can get worse and the game can become just unplayable. May UE5 be sunsetted. Symphony of War, the DLC. There is a decent variety of missions, though they do not continue the main story and are various side missions (happening before the main story). I guess, however cheesy the main story was, I did appreciate it existing. Still, the main squads can be deployed there, so it is a good way for the lower level ones to gain XP, as the missions are mostly easier (not all, but most so far). Some bits are odd, e.g. some characters are shown as knowing each other, but it was never shown in the base game.

I partially agree with your statement. The natural languages are heavily patterned, while most of the human knowledge is recorded in text, including the descriptions of the properties of various physical objects. The LLMs know that in the sense of having this data and building the relationships between various words, so they do have internal representations of concepts. They obviously cannot have it as personal physical experience in the same way as humans. However, the actions and feedback are included in the reinforcement learning and the user interactions (if the incorrect responses were rated higher than the correct ones, it can lead to issues). Programming languages and study materials tend to be more structured than random texts, so LLMs work better with them. You also do not need from your pair programmer whether they have a cold or if they are hungry or what they think about dogs (unless you are really bored). It's nice if they remind you to stay hydrated, but they do not need to experience it physically themselves (neither does a calendar reminder which is easier to set up). Therefore, they can be fit for the particular purpose. Well, given the wide adoption of Claude, they are. Here is to hope that the developers can understand the code they ship. However, as you've said, LLMs (and humans) can be wrong and cannot be held accountable for their errors (nor can some humans unless you consider Luigi Mangione to be inspirational, but, again, it's a more of a systemic social issue and not directly related to LLMs). Therefore, ideally you would not want either in the decision-making position. I agree that the tools should be fit for purpose and the job market can be challenging to navigate. I personally find filling the application forms with multiple popup menus on an external website to be more annoying, especially when the exact same information is in your CV and they cannot even scrape that correctly. One would hope that it'd discourage competitors, so the resulting pool is lower. Overall, the first rounds of interviews are to find the more suitable candidates and tend to be outsourced to the people who know little of the field you are to be working in. So, using a chatbot at this stage and just reading the summary or watching a video recording is not a bad idea. When you get to the point of the practical exercises and need to explain your logic, that's when you'd want your potential team lead to be present.

I would like to start with that the following are personal opinions, observations, and anecdotes and not a scientific study (alas, no data and not enough inclination for that). I can see the point in regard to the energy consumption and I also find it quite irritating that our social group, Gamers™, seek and encourage the higher use of energy on something as frivolous as graphical fluff (may UE5 be sunsetted). Regarding the necessity, it is very relative. I do not have mobility impairments and can use a broom instead of a vacuum cleaner (I believe the animal companion prefers less noise), and someone whose job and source of income is cleaning would go for the more "human energy"-efficient option. In the case of LLMs, a use case I've seen is job search, a very generic activity with a large amount of text on it. One of the people I know tried to use the free (government-funded) employment assistance services. The meatbags there were nigh useless and apparently could not parse the person's educational background and previous employment, while the positions suggested could have been just randomly pulled from a pool. On the other hand, the chatbot was able to provide the job titles for the desired career direction, what to watch out for in the adverts, how to format the CV, and how to pace the search, so it could be done alongside the ongoing employment without burning out. The LLM also was available at any time and provided responses and feedback promptly. Some people might prefer the LLMs as the pair programmer or a study partner for the same reasons - availability, flexibility, and general familiarity with the relevant field. Granted, they are/should be aware of the possibility of hallucinations and the necessity to check sources. Regarding taxation, at the moment, I would like to see how it is going to go. It is possible to tax the corporations (unlikely may it be) and the "agentic" AI is not able to do most jobs fully (even 2D artists'). And institutional knowledge is a thing that can easily get lost in the layoffs. So, I agree that the lack of employment due to the CEOs' lack of foresight and professional skills is a threat to the livelihoods of their employees and can negatively affect the companies and the end-users in the long run. The most recent case I am aware of is PinkNews going for "reporter-free newsroom" (the CEO is a dumb ****, so expected as much). So, the point being, there are areas where humans perform worse than the genAI, the necessity is relative, and the human CEOs not being concerned with the long-term prospects of their companies or the societal outcome of their decisions is an issue. The not mentioned issue with the LLMs and image generation being widely available is that malicious actors can use them as well, whether it is spear-phishing, various photo editing, or hate speech at scale. At what point an undesirable side effect becomes an inherent feature I cannot tell.

Symphony of War: The Nephilim Saga Chapter 12 I wish my ego was that big. He's not wrong. It is a pretty good trait. Chapter 17-19 (I think) RGB dragons. The water-walking clerics. The party has "interesting" ideas. The enemy is a significant chunk of the population. But the survivors will need less resources, so it evens out. Some of the later battles showed that the quality is more important that the quantity of the dragons. Best friends and roommates I could not tell if it was the writing or the translation (I believe the game was written in English). Thank the devs for the toggle The Sayunaa arc (spoilers) I.e. our power structures are for the betterment for the world. I guess, it is what a cleric would say. "Liberating" the other locations was fine, but the capital is a no-no. Chapter 27 or so Chapter 28 End-game spoilers The Legends DLC. I did bring more of my squads there, but the paladins could defeat the 20 hostile squads alone with enough time. Some of the unique mercenaries.

Possessor(s) Started moving before the room loaded. It was not, in fact, quite fine. The first attempt. The boss did not have the time to do it when I played more aggressively. He was not wrong. The first of the major bosses. I was sceptical. So, yeah. Mid-game spoiler The area boss

What I've purchased is mine to keep. The Black Geyser developers have launched a Kickstarter campaign for another RTwP RPG in the setting: https://www.kickstarter.com/projects/grapeocean/avaria-iron-rule The game is not related to Avaria: Chains of Lust.

It is the "garbage in, garbage out" situation - LLMs are tools and depend on the ones who make and use it, so Grok reflects Musk and the current Xitter population. Structured, high-quality data and reinforcement learning should provide better results, but require effort (time, funding, etc.). The energy consumption and e-waste are issues, though. Hence my general dislike of the corporate-owned models, despite their current affordability (also that the affordability can be easily taken away). One could consider the open-source software as an example of people cooperating, but I am unsure if it can work for the training and hosting. It is a good use-case. Alternatively, forcing companies to provide the contractual terms in the layperson-readable format could achieve the same result (some already do it). I know I am not paying a lawyer to check a random EULA for me, though I try to read them briefly (some are interesting, the software ones usually are "the software might not work, we will change it however we want, and you will not sue us").

I finished Possessor(s) a week ago and was quite happy with the game, some jank aside (mostly, the collectibles and the map). I liked the exploration and the boss battles, with the difficulty increasing as bosses were losing health. I have started Symphony of War: The Nephilim Saga. It is a tactical game in a fantasy setting. Unless there is a major plot twist at the end, it is fairly generic. The tactical combat and character development feel satisfying, with the "one more turn" lasting for 5 hours. The complexity is unbelievable for an RPGMaker game - there are squads, large maps, functional UI with mouse support. I think there is a perma death mode, but it is optional. The drawback is that the story and the narrative are hard to take seriously. The high-fantasy writing style is consistent - the generals and the farmers have a different manner of speech and the narrator leans towards the former, so it is not as obnoxious as it was with Gedonia. You can also just hire new units on the market, instead of waiting for your party members to multiply, which is most welcome (the last game of this type that I played is Fire Emblem: Fates).

The Life and Suffering of Prince Jerian has been funded successfully. https://www.kickstarter.com/projects/101xp/the-life-and-suffering-of-prince-jerian/posts/4665935

From quickly skimming through the article, it is solely about copyright rather than the environmental impact, employment, and the developers' ability to troubleshoot AI-generated code. As a man with an MSc in Data Science who does not work with genAI (not because I do not want to), I unironically love the topic. The 2D artists, whose livelihoods were affected by it, have significantly stronger feelings. Their logic is that whether or not you would have paid a person for the work, the models you've used likely were trained on their work. I would say that (I think) the labour for the sake of it, i.e. if it does not produce anything or gain skills (or at least some satisfaction or financial compensation), is soul-crushingly pointless and has no inherent value. There are several aspects which when combined might make one less comfortable. LLMs (Large Language Models) are trained to be extremely confident, yet supportive and go with the user's suggestions, because humans perceive confidence as knowledge (and there is some link between the eloquence and the perceived intelligence, which negatively affects primarily immigrants). LLMs are prediction models and do not possess any "ground truth", just a lot of data with different weights attached. They can work well for data summarisation or for some generic data, but less so for the niche subjects (and if you are unfamiliar with the field, you might not be able to spot errors). At the moment, the older models are provided for free to build reliance on them, as the skills unused deteriorate, so it is expected to lead to dependence (I can tell that I cannot easily multiply 3+ digit numbers without writing them down). You can see the similar pattern (en****tification) of building a user base, then extracting value from it in the other industries, such as video streaming. For software development in particular, I've been told that Claude Opus is a fantastic tool to use. The gotcha being that the developers must understand and be able to troubleshoot the code it generates, otherwise, the software will be impossible to support long-term. There are some other drawbacks and use cases and most are summarised in Abigail Thorn's video (1h): https://www.youtube.com/watch?v=AaU6tI2pb3M If you do need an LLM in your life, I would still suggest running one locally (can be done with a £700 Mac Mini M2).

Possessor(s) is available on GOG with a 75% discount. I've repurchased and left a longer review there. TL;DR - a fantastic game with excellent story and combat, and somewhat unfriendly map. https://www.gog.com/en/game/possessors

If you are talking about Jerian, then it is a stats-heavy visual novel with a poor UI. If you like the particular genre, I'd suggest Long Live the Queen instead. One of the dev logs was amusing for the wrong reasons: https://steamcommunity.com/games/2936290/announcements/detail/529872149958100080?snr=2___ Like, the lack of self-awareness is hilarious - the developers are the ones in complete control of the setting, of the choices available and their consequences. The refusal to acknowledge the base assumptions and biases can and will bleed in, whether in the selection of options or in the wording. Due to the amount of variations, the narrative can be collaborative, but, ultimately, the developers are the one determining the limits of what is possible. Well, also the modders, but they rarely work with visual novels. The Prince could make a ground-breaking industrial invention and realise that he wants to implement universal basic income or he could spend the whole game trying to gather the ultimate collection of post stamps. I can tell that neither option will be offered despite them being well within the setting. I have backed it, as a family member wants the game, but still find it unappealing personally.

Possessor(s). I love the combat system and the passive that sends the projectiles right back at the attacker. After I found where to go (the map did not cooperate), the story picked up the pace and progresses steadily - there's a decent character development and lore (alas, no lore dumps). Judging by the achievements and an in-game side quest, there are more powerful versions of the main bosses. The one I have fought had a completely different moveset to its main story counterpart. It also was sitting on one of the main weapons (I've gathered all of them by now and fully upgraded 2). There are some rough edges, I assume due to the time constraints (the exploration dialogues seem to be out of sequence), but it is very enjoyable overall.

The Fortune's Run developer has been released and has plans to release the game (as in, it still brings her some income and personal fulfilment to do so). https://steamcommunity.com/games/1692240/announcements/detail/492721254199986236?snr=2___ I bought it on Steam some time ago, though have not played yet. The game is available on GOG as well.